EU’s AI Act
Seoul needs prudent steps to address possible dangers
The European Parliament passed the AI Act on Wednesday, which is the world’s first comprehensive statute designed to address the potential dangers of artificial intelligence (AI). The passage came five years after the introduction of a relevant bill. At first, the legislative process hit a snag due to differing stances held by the 27 member countries of the European Union.
As the rapid advancement of generative AI fueled a sense of urgency, discussions gained momentum. The resounding endorsement of the AI Act signifies the collective acknowledgment among EU members regarding the urgent need to address the potential risks linked with AI.
In a related context, President Yoon Suk Yeol voiced caution on Monday regarding the challenges AI-generated fake news poses to democratic societies.
“Fake news and false information [made with] AI and digital technology are infringing upon individual freedom and human rights while threatening the base of democracy,” Yoon said during the opening of the ministerial meeting of the 3rd Summit for Democracy held at The Shilla Seoul hotel.
The EU’s act is scheduled for phased implementation from the second half of this year. Enterprises involved in producing highrisk technologies, such as autonomous driving and medical equipment, are mandated to disclose pertinent data paired with strict pretests. Furthermore, the act specifies that individuals seeking to deploy high-risk technologies must engage human supervisors in the process.
Those trying to use deepfake images or videos also need to clarify that they are generated by AI. Except in cases of investigating serious crimes, the utilization of real-time biological recognition systems will be prohibited. Any business found violating these regulations will face significant penalties amounting to 7 percent of their annual sales or up to 35 million euros ($38 million).
Concerns have grown regarding the potentially perilous outcomes that AI can cause. Recently, anxiety has escalated surrounding the potential hazards of fabricating facts using generative AI. There have been many cases of stock price manipulation and even election outcomes affected by fake images and videos made by AI.
It is becoming increasingly apparent that unless AI development is steered in a human-centric direction through regulation, humanity could encounter catastrophic consequences. Failure to properly regulate AI and artificial general intelligence (AGI) under appropriate guidelines could lead to a dire scenario. AGI, which surpasses human capabilities, poses a particularly significant risk if left unchecked.
It is natural for governments around the world to go all out in preparing measures to cope with the potential damage that AI can cause.
In a similar vein, the Yoon Suk Yeol administration should marshal all available resources to regulate AI effectively and mitigate potential adverse effects. Korea took the initiative by introducing the AI Basic Act early last year, encompassing fundamental concepts, industry promotion, and regulation. However, the pertinent bill has languished in the National Assembly, showing no indications of advancement as rival parties focus on preparations for the upcoming April 10 general elections.
We need to focus on introducing appropriate standards that can minimize the adverse impact of AI. Using AI properly will guarantee an innovative, bright future for humans. We must adopt measures to ensure sustained prosperity facilitated by advanced technologies. However, we must also exercise caution in implementing overly stringent regulations, as this could potentially hinder Korea’s competitiveness in the global AI race.
Global AI giants are rushing to make huge investments in the AI sector. Seoul has a long way to catch up with the advanced countries in state-of-the-art AI technology. Now is the time to positively consider audacious assistance to AI companies, while cautiously approaching ethical problems that can be triggered by the technology.